A method for regulating energy storage graded charging and discharging for photovoltaic power fluctuation smoothing

Abstract

This invention discloses a graded charging and discharging regulation method for photovoltaic power fluctuation mitigation, relating to the field of photovoltaic power generation and energy storage coordinated control technology. This method is applied to photovoltaic power plant systems comprising string-level self-balancing grid units, station-level central grid-connected energy storage systems, and station-level coordinated grid controllers. Through real-time monitoring and compensation of power deviations caused by local shading by string-level units, and by performing coordinated power scheduling and dynamic reactive power allocation at the station level, it achieves coordinated optimization of power balance within the photovoltaic array and grid support at the grid connection point. Simultaneously, through virtual synchronous machine cluster control and protective off-grid / re-grid mechanisms, it enhances the inertial response capability and fault ride-through performance of the power plant. This invention can effectively mitigate power fluctuations in complex shading environments, improve the power generation efficiency and grid interaction capability of photovoltaic power plants, and enhance the stability and reliability of the power system.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic power generation and energy storage synergistic control technology, specifically a method for graded charging and discharging regulation of energy storage to smooth out photovoltaic power fluctuations. Background Technology

[0002] With the global energy structure shifting towards cleaner energy sources, the installed capacity and penetration rate of photovoltaic (PV) power generation are rapidly increasing. Large-scale centralized PV power plants are typically deployed in open areas, making their power output susceptible to environmental factors such as cloud drift, localized dust accumulation, or shading, resulting in significant randomness and volatility. This volatility not only affects the economic output of the power plant itself, but when connected to the grid via power electronic equipment, the power disturbances, lacking inertia and damping, are transmitted to the grid. Particularly in areas with relatively weak grid structures, this can lead to grid frequency shifts, voltage flicker, and even stability issues, hindering the high-proportion safe absorption of PV power. Therefore, effectively mitigating PV power fluctuations and enhancing the active support capabilities of PV power plants for the grid have become urgent technical challenges in this field.

[0003] To address these issues, existing technologies have proposed various power mitigation and grid support solutions combining energy storage systems. One type of solution focuses on power optimization within the photovoltaic array, compensating for the current of shaded components to ensure consistent current in series branches, thereby increasing the array's maximum output power under shading conditions. While this approach improves local power generation efficiency, its energy storage unit functions solely as a current compensation source, lacking interaction with the grid and the ability to mitigate grid-connected power fluctuations or provide grid support. Another type of solution emphasizes overall control of the power plant's grid connection point, simulating the external characteristics of a synchronous generator to provide voltage, frequency, and inertia support to the grid. However, this approach treats the photovoltaic power plant as a single power source, lacking the ability to perceive and effectively manage the severe power imbalances within the array's branches caused by localized shading. When severe power mismatch exists within the array, the overall output characteristics of the power plant deteriorate, significantly increasing the mitigation burden on the station-level energy storage and even affecting its grid support effectiveness and reliability, failing to address the problem at its root.

[0004] In summary, existing technical solutions for addressing localized shading issues either focus solely on power generation optimization within the photovoltaic array, failing to extend to grid support, or adopt a broad approach to power plant-wide balancing and grid construction, neglecting the refined handling of the root cause of internal power mismatch. This disconnect in technical pathways makes it difficult to simultaneously optimize photovoltaic array power and strengthen grid connection at the grid connection point under complex shading conditions. Therefore, an innovative control method is needed to deeply integrate and coordinate the refined balancing compensation within the photovoltaic array with the power plant's proactive grid support within a unified framework.

[0005] The purpose of this invention is to overcome the limitations of the prior art and provide a method for graded charging and discharging regulation of energy storage that can simultaneously solve the dual problems of internal power mismatch and external power grid support. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a graded charging and discharging regulation method for photovoltaic power fluctuation smoothing. Through a multi-level collaborative mechanism of string-level real-time detection and compensation, station-level collaborative scheduling and reactive power support, and virtual synchronous machine cluster control, it can effectively address the power mismatch problem caused by local shading within the photovoltaic array and significantly improve the photovoltaic power station's active support capability for grid voltage, frequency, and inertia, thereby smoothing power fluctuations at the source and enhancing system stability at the grid.

[0007] To solve the above-mentioned technical problems, this invention provides the following technical solution: a photovoltaic power station system with energy storage-based graded charge and discharge regulation for smoothing photovoltaic power fluctuations, applicable to a photovoltaic power station system comprising multiple photovoltaic strings, string-level self-balancing grid units corresponding to each photovoltaic string, a station-level central grid-connected energy storage system, and a station-level collaborative grid controller. The method includes the following steps:

[0008] Step 1: Each string-level self-balancing grid unit monitors the actual output power of its corresponding photovoltaic string in real time, and calculates the theoretical output power of the photovoltaic string under the current environment based on the operating environment parameters of the photovoltaic string. The string-level self-balancing grid unit includes an energy storage battery and a bidirectional converter with independent grid control capability.

[0009] Step 2: Each string-level self-balancing grid unit calculates the power difference between the actual output power and the theoretical output power, and compares the power difference with a preset first power threshold. When the power difference is greater than the first power threshold, the photovoltaic string is determined to be a weak string affected by local shading, and the local compensation mode is activated.

[0010] Step 3: The string-level self-balancing grid unit in the local compensation mode controls the energy storage battery to discharge through the bidirectional converter, so that the total active power output by the string-level self-balancing grid unit approaches the theoretical output power. At the same time, the bidirectional converter of the string-level self-balancing grid unit operates in grid mode to maintain the voltage and frequency stability of the string AC bus connected to the string-level self-balancing grid unit.

[0011] Step four: The station-level collaborative grid controller collects the operating status, power difference, and state of charge information of all string-level self-balancing grid units through the communication network, and monitors the voltage and frequency of the photovoltaic power station grid connection point.

[0012] Step 5: When the state of charge (SOC) of the energy storage battery of any string-level self-balancing grid unit in local compensation mode is lower than the preset SOC threshold, the station-level collaborative grid controller initiates power collaborative scheduling. The power collaborative scheduling includes instructing the station-level central grid energy storage system to inject power into the string AC bus, or instructing other string-level self-balancing grid units with an SOC higher than the SOC threshold to adjust the output power of the other string-level self-balancing grid units, so as to provide power support to the string-level self-balancing grid unit with the low SOC of the energy storage battery through the string AC bus.

[0013] Step six: When the voltage at the grid connection point deviates from the reference voltage value set for that grid connection point, the station-level collaborative grid controller calculates the total reactive power required for the entire station based on the voltage deviation, and dynamically distributes reactive power adjustment commands to all string-level self-balancing grid units and the station-level central grid energy storage system. This enables each string-level self-balancing grid unit to coordinate the generation or absorption of reactive power while completing its own active power adjustment tasks, thereby supporting the stability of the grid connection point voltage.

[0014] Furthermore, in step two, the algorithms for calculating the power difference and determining the weaker strings are executed by the local controller embedded in each string-level self-balancing network unit. The theoretical output power... The calculation formula is: ;

[0015] in, This represents the theoretical output power of the photovoltaic string under the current operating environment. This indicates the photoelectric conversion efficiency of a photovoltaic module under standard test conditions. This represents the total area of ​​all photovoltaic modules in the photovoltaic string. This represents the solar irradiance collected in real time by the irradiance sensor installed on the string. This indicates the real-time temperature of the photovoltaic module's backsheet, as collected by a temperature sensor. This indicates the reference temperature under standard test conditions. This represents the power temperature coefficient of a photovoltaic module.

[0016] Furthermore, in step three, the string-level self-balancing grid-building unit operates in grid-building mode as follows: the bidirectional converter adopts a voltage source control strategy. The bidirectional converter adjusts the modulation amplitude and phase through its own controller to autonomously establish and maintain the voltage amplitude and frequency reference of the string AC bus, providing voltage and frequency support for other power electronic devices connected to the same bus, without relying on the voltage and frequency signals of the external power grid.

[0017] Furthermore, the power coordinated scheduling in step five specifically includes: the station-level coordinated network controller establishing a scheduling model with the optimization objectives of overall station power balance and balanced state of charge of each energy storage unit, and solving for the optimal power command of each controllable unit. The objective function of the scheduling model is based on minimizing the output power change rate of the station-level central energy storage and maximizing the consistency of the state of charge of each energy storage unit.

[0018] Furthermore, in step six, the specific method for the station-level collaborative grid controller to dynamically allocate reactive power adjustment commands is as follows: based on the current apparent power capacity margin and response priority coefficient of each grid unit, namely each group of cascade self-balancing grid units and the station-level central grid energy storage system, the total reactive power demand is allocated proportionally. Among them, units with larger capacity margins are allocated larger reactive power adjustment amounts, and units in idle or light-load states have higher response priorities.

[0019] Furthermore, the method also includes:

[0020] Step 7, Virtual Synchronous Machine Cluster Control: The station-level collaborative grid controller uniformly configures virtual rotational inertia parameters and virtual damping parameters for all string-level self-balancing grid units and the bidirectional converters of the station-level central grid energy storage system. When these bidirectional converters respond to changes in grid frequency, they simulate the rotor motion equations of synchronous generators, providing inertial support and damping for the grid. The frequency-active power droop characteristic equation implemented by the virtual synchronous machine control algorithm is as follows: ;

[0021] in, This indicates the active power output value that the network unit needs to adjust based on the frequency deviation. Indicates the frequency droop factor. Indicates the reference value of the rated frequency. This represents the actual frequency of the power grid as collected in real time. This represents the set virtual moment of inertia.

[0022] Furthermore, in step seven, all grid-building units, under the unified coordination of the station-level collaborative grid-building controller, adopt consistent virtual rotational inertia parameters and virtual damping parameters, so that the entire photovoltaic power station appears to the outside world as an equivalent virtual synchronous generator with uniform inertia and damping characteristics, participating in the primary frequency regulation of the power grid.

[0023] Furthermore, the string-level self-balancing network building unit communicates with the station-level collaborative network building controller, as well as with each string-level self-balancing network building unit, via high-speed industrial Ethernet based on time-sensitive networking technology. This ensures that power difference information, scheduling instructions, and status feedback information are transmitted within milliseconds to meet the stringent requirements of rapid collaborative control for communication latency.

[0024] Furthermore, the energy storage battery is a lithium iron phosphate battery or a lithium titanate battery, the bidirectional converter is a three-phase voltage source converter with a full-bridge or three-level topology, and the switching device used in the bidirectional converter is an insulated-gate bipolar transistor or a silicon carbide metal oxide semiconductor field-effect transistor.

[0025] Furthermore, the method also includes:

[0026] Step 8, Protective Off-Grid and Seamless Reconnection: When a serious fault is detected in the string AC bus or the main power grid of the power station, the station-level collaborative grid controller instructs all grid units in the affected area to switch to islanded operation mode. The grid units in this area continue to work together to maintain the voltage and frequency stability of the local power grid. After the fault is cleared and the power grid voltage and frequency return to normal, the station-level collaborative grid controller adjusts the output voltage phase and amplitude of each grid unit to synchronize these voltages with the main power grid, achieving smooth, shock-free grid connection of all units.

[0027] Compared with existing technologies, this method for graded charge and discharge regulation of photovoltaic power fluctuation smoothing has the following advantages:

[0028] I. This invention uses string-level self-balancing grid units to detect and compensate for power loss caused by local shading in real time, and combines a station-level collaborative grid controller to perform power collaborative scheduling and dynamic reactive power allocation. This enables both fine-grained power balancing within the photovoltaic array and active grid support at the grid connection point, thereby effectively mitigating photovoltaic power fluctuations and improving the overall power generation efficiency and grid stability of photovoltaic power plants under complex shading conditions.

[0029] Second, by uniformly configuring virtual synchronous machine control parameters for all grid-connected units and quickly switching to islanded operation mode and achieving seamless reconnection in case of faults, this invention enables the photovoltaic power station as a whole to have inertia, damping and autonomous grid-connection capabilities similar to a synchronous generator, thereby enhancing the frequency regulation and fault ride-through capability of the power grid and improving the operational reliability and security of the power system in scenarios with a high proportion of photovoltaic access.

[0030] Other advantages, objectives and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination or study, or may be learned from the practice of the invention. Attached Figure Description

[0031] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without any creative effort.

[0032] Figure 1 This is a diagram illustrating the method steps of the present invention;

[0033] Figure 2 This is a schematic diagram of the system architecture and energy / information flow of the present invention;

[0034] Figure 3 This is a flowchart of the core logic of hierarchical regulation in this invention. Detailed Implementation

[0035] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0036] Example

[0037] like Figures 1 to 3 As shown in this embodiment, the photovoltaic power fluctuation smoothing energy storage hierarchical charging and discharging regulation method disclosed is applied to a large-scale centralized photovoltaic power station system. The system specifically includes 30 photovoltaic strings, 30 string-level self-balancing grid-building units corresponding to each photovoltaic string, a station-level central grid-building energy storage system, and a station-level collaborative grid-building controller. All devices are connected via high-speed industrial Ethernet based on time-sensitive networking technology to ensure rapid transmission of control commands and status information.

[0038] Photovoltaic string: Each photovoltaic string consists of 22 monocrystalline silicon photovoltaic modules connected in series. The rated power of a single photovoltaic module is 550W, the open-circuit voltage is 45V, and the short-circuit current is 13.5A. The photovoltaic string is installed at a tilt angle of 30° with the azimuth facing due south, and is used to convert solar energy into electrical energy output.

[0039] String-level self-balancing grid unit: Each string-level self-balancing grid unit integrates an energy storage battery and a bidirectional converter. The energy storage battery uses lithium iron phosphate batteries with a rated capacity of 10kWh, a rated voltage of 51.2V, a charge / discharge rate of 1C, and a state-of-charge (SOC) range of 10% to 90%. The bidirectional converter is a three-phase voltage source converter with a three-level topology, a rated power of 12kW, an input voltage range of 45V to 60V, an output voltage of 380V, and an output frequency of 50Hz. The switching devices are insulated-gate bipolar transistors with a switching frequency of 10kHz, providing independent grid control capabilities and bidirectional power transmission. In addition, each string-level self-balancing grid unit also embeds a local controller using a high-performance STM32H743 microprocessor for functions such as power monitoring, theoretical power calculation, weak string identification, and local compensation control.

[0040] Station-level central grid-connected energy storage system: The station-level central grid-connected energy storage system consists of lithium iron phosphate battery clusters, bidirectional converter cabinets, and local monitoring units. The total rated capacity is 500kWh, the rated power is 250kW, the rated voltage of the energy storage battery clusters is 600V, the charge / discharge rate is 1C, and the state of charge operation range is 15% to 85%. The bidirectional converter cabinet adopts a three-phase voltage source converter with a full-bridge topology and an output voltage of 10kV. It is connected to the main bus of the photovoltaic power station through a transformer and has independent regulation capabilities for active and reactive power.

[0041] Station-level collaborative network controller: The station-level collaborative network controller adopts an industrial-grade server equipped with an Intel Core i7 processor, 32GB of memory, 1TB of storage capacity, and runs a real-time operating system. This controller establishes communication with all string-level self-balancing network units and the station-level central network energy storage system via high-speed industrial Ethernet. The communication rate is 1Gbps, and the transmission latency is ≤10ms. It is used to collect operating status information from each unit, perform power collaborative scheduling, reactive power allocation, and virtual synchronous machine cluster control, among other functions.

[0042] Auxiliary testing equipment: Each photovoltaic string is equipped with one irradiance sensor and one temperature sensor. The irradiance sensor is a TBQ-2 model, with a measurement range of 0 to 2000 W / m² and a measurement accuracy of ±5%. The temperature sensor is a PT100 platinum resistance thermometer, with a measurement range of -40℃ to 120℃ and a measurement accuracy of ±0.5℃, used to collect real-time solar irradiance and photovoltaic module backsheet temperature. Voltage transformers and current transformers are installed at the grid connection point of the photovoltaic power station. The voltage transformer ratio is 10kV / 100V, and the current transformer ratio is 200A / 5A, used to monitor the voltage and current signals at the grid connection point.

[0043] In this embodiment, the method steps are implemented in detail as follows:

[0044] Step 1: Actual output power monitoring and theoretical output power calculation:

[0045] Each photovoltaic (PV) string-level self-balancing grid unit's local controller, through its internal voltage and current sampling modules, collects the PV string's output voltage and current in real time at a sampling frequency of 1kHz. Based on the active power calculation formula... The actual output power of the photovoltaic string is calculated. ,in This represents the effective value of the output line voltage of the photovoltaic string obtained from sampling. This represents the effective value of the output line current of the photovoltaic string obtained from sampling. The power factor is obtained through phase-locked loop (PLL) technology and is measured during normal operation. The value ranges from 0.98 to 1.0.

[0046] Meanwhile, the local controller calculates the theoretical output power under the current environment based on the operating environment parameters of the photovoltaic string using the theoretical output power formula. The formula is as follows: ;

[0047] Detailed explanations of each parameter are as follows:

[0048] The photoelectric conversion efficiency of the photovoltaic module under standard test conditions, defined as irradiance of 1000 W / m², module temperature of 25°C, and atmospheric quality AM1.5. The monocrystalline silicon photovoltaic module used in this embodiment... The value is 23.5%. This parameter is the calibration value when the photovoltaic module leaves the factory and can be obtained by querying the module datasheet.

[0049] The total area of ​​all photovoltaic modules in the photovoltaic string. In this embodiment, each photovoltaic module measures 1.6m × 1.0m, with a single module area of ​​1.6m². Each photovoltaic string contains 22 modules. Therefore... The unit is square meters.

[0050] The solar irradiance is collected in real time by an irradiance sensor installed on the string, with the unit being watts per square meter. The sensor outputs the collected data every 100ms. The local controller performs a moving average filter on the collected data to remove random interference. The filter window length is 10 sampling points.

[0051] The temperature of the photovoltaic module backsheet is collected in real time by a temperature sensor, in degrees Celsius. The sensor outputs the collected data every 100ms. The local controller performs first-order low-pass filtering on the collected data, with a filtering time constant of 0.1s.

[0052] The reference temperature under standard test conditions is 25℃. This parameter is a recognized standard test condition parameter in the photovoltaic industry and is used to unify the temperature benchmark for photovoltaic module performance calibration.

[0053] The power temperature coefficient of a photovoltaic module, expressed as % / ℃, represents the percentage decrease in output power for every 1℃ increase in module temperature. In this embodiment, a monocrystalline silicon photovoltaic module is used. The value is -0.38% / ℃. This parameter is the factory calibration value of the component and can be obtained by looking up the component datasheet. The negative sign indicates that the output power decreases when the temperature increases.

[0054] Example calculation: Suppose that at a certain moment, the irradiance sensor collects... Temperature sensor data , Then substitute into the formula to calculate: ;

[0055] Step 2, Power Difference Calculation and Weak String Identification:

[0056] The local controller of each cascaded self-balancing network unit determines the actual output power obtained in step one. and theoretical output power Calculate the power difference The unit is watts.

[0057] The local controller presets a first power threshold. The threshold is set based on the rated power of the photovoltaic string and the allowable range of power fluctuations during actual operation. In this embodiment, the rated power of the photovoltaic string is 12.1kW, with 22 strings × 550W. Considering the power deviation caused by normal environmental fluctuations, the threshold is set... That is, 10% of the rated power.

[0058] The local controller will calculate the power difference. With the first power threshold Comparison:

[0059] when At that time, it was determined that the photovoltaic string was operating normally, with no obvious local shading effect, the string-level self-balancing grid unit maintained the normal operating mode, and the energy storage battery did not perform charging and discharging compensation.

[0060] when When the photovoltaic string is determined to be a weak string affected by local shading, the local controller activates the local compensation mode and uploads the activation status information to the station-level collaborative network controller via the communication network.

[0061] Example: Suppose a photovoltaic string has... The actual data collected and calculated ,but The string is determined to be a weak string, and the local compensation mode is activated.

[0062] Step 3: Local compensation mode operation and string AC bus stability control:

[0063] In string-level self-balancing grid units operating in local compensation mode, the local controller adjusts the power difference based on the local compensation mode. Generate discharge power command for energy storage battery Ignoring converter losses, in practical applications, losses can be corrected using a loss compensation coefficient. This allows the energy storage battery to discharge to the string AC bus via a bidirectional converter, thereby increasing the total active power output of the string-level self-balancing grid unit. Approaching the theoretical output power .

[0064] Meanwhile, the bidirectional converter of the string-level self-balancing grid unit operates in grid mode, specifically adopting a voltage source control strategy. Its core is to adjust the amplitude and phase of the modulation wave through the controller of the bidirectional converter, and autonomously establish and maintain the voltage amplitude and frequency reference of the string AC bus.

[0065] Specific implementation of the voltage source control strategy: The controller of the bidirectional converter adopts a droop control algorithm, and the voltage amplitude droop characteristic is as follows: The frequency droop characteristic is ,in:

[0066] The rated voltage amplitude of the string AC bus is set to 380V in this embodiment, which is the effective value of the line voltage.

[0067] This is the voltage droop factor, with a value of 0.001V / Var, in volts per var, used to adjust the voltage amplitude according to the reactive power output.

[0068] This represents the reactive power output value of the bidirectional converter, expressed in volt-amperes (VA).

[0069] This is the reference value for the rated frequency, which is 50Hz.

[0070] This is the frequency droop factor, with a value of 0.002 Hz / kW, in Hertz per kilowatt, used to adjust the frequency based on active power output.

[0071] The bidirectional converter acquires the voltage and frequency signals of the string AC bus in real time through its own controller, compares them with a reference value, and generates amplitude and phase control signals for the modulation wave through a PI regulator. This drives the insulated-gate bipolar transistor (IGBT) to switch, achieving closed-loop control of the output voltage and frequency. This grid configuration does not rely on the voltage and frequency signals of the external power grid and can autonomously establish a stable string AC bus voltage and frequency reference. It provides reliable voltage and frequency support for other power electronic devices connected to the same bus, ensuring the stable operation of the power system at the string level.

[0072] Example: A weak string The local controller instructs the energy storage battery to discharge at a power of 1.78kW. After conversion by the bidirectional converter, the output is sent to the string AC bus. At this time, the total active power of the string-level self-balancing grid unit is... The output power is consistent with the theoretical output power. The bidirectional converter stabilizes the string AC bus voltage within the range of 380V±5V and the frequency within the range of 50Hz±0.2Hz through a voltage source control strategy.

[0073] Step 4: Collection of operational status information and monitoring of grid connection points:

[0074] The station-level collaborative network controller periodically collects the operating status information of all string-level self-balancing network units through a high-speed industrial Ethernet based on time-sensitive networking technology, according to a preset communication protocol, with a collection period of 100ms.

[0075] The collected information includes the following:

[0076] Operating status: Normal operating mode or local compensation mode.

[0077] Power parameters: actual output power Theoretical output power Power difference .

[0078] Energy storage battery status includes: State of Charge (SOC), battery voltage, and battery temperature. The SOC is calculated using the ampere-hour integration method, with the following formula: , The initial state of charge, For battery charging and discharging current, In this embodiment, the rated capacity of the battery is... .

[0079] Meanwhile, the station-level collaborative grid controller collects voltage and current signals through voltage and current transformers installed at the grid connection point, with a sampling frequency of 1kHz, and calculates the grid connection point voltage. and frequency And monitor its changes in real time.

[0080] Step 5, Power Coordination Scheduling Execution:

[0081] The station-level collaborative grid controller monitors the state of charge (SOC) of the energy storage batteries of each string-level self-balancing grid unit in local compensation mode in real time, and compares it with a preset SOC threshold. A comparison is made. In this embodiment, considering the cycle life and depth of discharge limitations of the energy storage battery, the following settings are made: .

[0082] When the SOC of the energy storage battery of any string-level self-balancing grid unit in local compensation mode is lower than At this time, the station-level collaborative network controller initiates power collaborative scheduling, and the specific process is as follows:

[0083] Scheduling model establishment: The station-level collaborative network controller establishes a scheduling model with the optimization objectives of overall station power balance and balanced state of charge of each energy storage unit.

[0084] The objective function is: ;

[0085] in, , The weighting coefficient is set according to actual operational requirements. In this embodiment... , , dimensionless.

[0086] This represents the change in output power of the station-level central grid energy storage system at time k, expressed in kW.

[0087] In this embodiment, the number of string-level self-balancing network units is [number to be specified]. .

[0088] Let represent the state of charge of the i-th string-level self-balancing grid unit energy storage battery.

[0089] This represents the average state of charge (SOC) of all self-balancing grid-connected energy storage cells in the string-level unit, expressed in units of %.

[0090] The constraints include:

[0091] Energy storage battery charge and discharge power constraints: In this embodiment .

[0092] Energy storage battery state of charge constraints: In this embodiment , .

[0093] Power constraints of station-level central grid-connected energy storage systems: In this embodiment .

[0094] Optimal power command solution: The scheduling model is solved using a particle swarm optimization algorithm to obtain the optimal power command for each controllable unit. The parameters of the particle swarm optimization algorithm are set as follows: number of particles 50, number of iterations 30, inertia weight 0.7, cognitive factor 1.5, social factor 1.5, and solution accuracy 0.1kW.

[0095] Power support execution: When the station-level central grid-connected energy storage system has power support capability, the station-level collaborative grid-connected controller issues a power injection command to it. It controls the injection of active power into the string AC bus through the bidirectional converter. This power is then transmitted through the string AC bus to the string-level self-balancing grid unit where the energy storage battery SOC is low, to charge the energy storage battery or directly supplement its output power gap.

[0096] When the state of charge is higher than When constructing other string-level self-balancing network units, the station-level collaborative network controller issues power adjustment commands to these units. Adjust its output power to provide power support to units with low SOC through the string AC bus.

[0097] Example: Suppose that the SOC of the energy storage battery in a string-level self-balancing grid unit in local compensation mode drops to 18%, and the power difference... The station-level collaborative grid controller, through a scheduling model, obtains an injection power command of 1.8kW for the station-level central grid-connected energy storage system. Simultaneously, it instructs two adjacent string-level self-balancing grid-connected units with a state of charge (SOC) of 60% to each adjust their output power by 0.1kW. The station-level central grid-connected energy storage system injects 1.8kW of active power into the string's AC bus according to the command, supplementing the power gap of the weak string. At this point, the discharge power of the energy storage battery in that string decreases, the SOC stops declining continuously, and gradually stabilizes.

[0098] Step Six: Dynamic Reactive Power Distribution and Grid Connection Point Voltage Support:

[0099] The station-level collaborative grid controller monitors the voltage at the grid connection point in real time. and the preset grid connection point reference voltage (In this embodiment, the voltage is set to 10kV, the effective value of the line voltage) The voltage deviation is calculated by comparing the values. The unit is kilovolts.

[0100] when That is, when the voltage deviation exceeds 3%, the station-level collaborative grid controller calculates the total reactive power required by the entire station based on the voltage deviation. The calculation uses a proportional-integral control algorithm: ;

[0101] in This is the proportionality coefficient. The integral coefficient is determined through engineering debugging to ensure the speed and stability of reactive power regulation.

[0102] The station-level collaborative network controller determines the current apparent power capacity margin of each network unit. and response priority coefficient The total reactive power demand is allocated proportionally, and the specific allocation formula is as follows: ;

[0103] The parameters are explained below:

[0104] Apparent power capacity margin The unit is kilovolt-amperes; among which Let be the rated apparent power of the i-th grid cell. This represents the current active power output value. This represents the current reactive power output value.

[0105] Response Priority Coefficient Setting rules: Network elements in an idle state, Network elements under light load conditions Mesh elements under heavy load 3.

[0106] The station-level collaborative grid controller sends reactive power adjustment commands to each grid unit through the communication network. Each cascade self-balancing grid unit and the bidirectional converter of the station-level central grid energy storage system, according to the commands, coordinates the generation or absorption of reactive power while completing their respective active power adjustment tasks, so as to support the voltage stability of the grid connection point.

[0107] Example: Assume the grid connection point reference voltage Actual monitoring ,but Calculate the total reactive power demand: (The integral term is taken as 2s in steady state, i.e.) ),but .

[0108] Assume there are 10 network units in the system, of which 3 are idle ( , ), 5 light-load ( , ), 2 heavy-duty ( , Then the total weight sum is:

[0109] ;

[0110] The reactive power allocation of a certain idle grid unit is:

[0111] ;

[0112] The grid unit sends out 18.5kVar of reactive power according to the instruction, and other grid units allocate reactive power adjustment tasks proportionally to support the grid connection point voltage to recover to the range of 10kV±0.3kV.

[0113] Step 7, Virtual Synchronizer Cluster Control Execution:

[0114] The station-level collaborative grid controller is a bidirectional converter for all string-level self-balancing grid units and the station-level central grid energy storage system, with unified configuration of virtual rotational inertia parameters. and virtual damping parameters (In this embodiment) This parameter configuration is sent to the controller of each network unit through the communication network and stored in the local register.

[0115] When the grid frequency changes, the bidirectional converters of each grid unit simulate the rotor motion equations of a synchronous generator according to the virtual synchronous machine control algorithm, providing inertial support and damping for the grid. The frequency-active power droop characteristic equation implemented by the virtual synchronous machine control algorithm is as follows:

[0116] ;

[0117] Detailed explanations of each parameter are as follows:

[0118] The active power output value that the network unit needs to adjust according to the frequency deviation, in kilowatts. A positive value indicates an increase in active power output, and a negative value indicates a decrease in active power output.

[0119] Frequency droop factor, in kilowatts per hertz, is set to 50 kW / Hz in this embodiment and is used to adjust the active power according to the static frequency deviation.

[0120] : Rated frequency reference value, taken as 50Hz, unit is Hertz.

[0121] The actual frequency of the power grid, measured in Hertz, is collected in real time and obtained through the frequency monitoring module at the grid connection point.

[0122] The virtual moment of inertia is set in kilograms per square meter, simulating the moment of inertia of the synchronous generator rotor. Increasing this parameter can improve the frequency stability of the system, but will reduce the frequency response speed; decreasing it will have the opposite effect. The value in this embodiment was determined through engineering experiments.

[0123] The rate of change of frequency deviation, measured in Hertz per second, is obtained by differentiating the frequency deviation signal using a first-order difference algorithm.

[0124] Virtual damping parameter, in kilowatts per hertz, is used to suppress frequency oscillations and improve the dynamic stability of the system. The value in this embodiment was determined through engineering debugging.

[0125] Under the unified coordination of the station-level collaborative grid controller, all grid-connecting units adopt consistent virtual rotational inertia parameters and virtual damping parameters, so that the entire photovoltaic power station appears to the outside world as an equivalent virtual synchronous generator with uniform inertia and damping characteristics, participating in the primary frequency regulation of the power grid.

[0126] Example: Assume the actual frequency of the power grid Frequency deviation Frequency deviation change rate ,but:

[0127]

[0128] Each grid unit increases its active power output according to its own power allocation ratio, thereby increasing the total active power of the entire photovoltaic power station, which suppresses further decline in grid frequency and supports grid frequency stability.

[0129] Step 8: Implementation of protective off-grid and seamless reconnection:

[0130] The station-level collaborative network controller monitors the operating status of the string AC bus and the main power grid of the power station in real time. It determines whether a serious fault has occurred by using the characteristic quantities of voltage and current signals. The criteria for determining a serious fault are as follows:

[0131] The string AC bus voltage drops to below 50% of the rated value for more than 10ms.

[0132] The voltage of the power plant's main grid drops to below 60% of its rated value and lasts for more than 10ms.

[0133] The power grid frequency deviation exceeds ±2Hz and lasts for more than 5ms.

[0134] Upon detecting the aforementioned severe fault, the station-level collaborative network controller immediately issues an islanding operation command via the communication network, instructing all network units within the affected area to switch to islanding operation mode. In islanding operation mode, each network unit continues to employ a voltage source control strategy to collaboratively maintain the voltage and frequency stability of the local power grid. The voltage is stabilized within the range of 380V±10V, and the frequency is stabilized within the range of 50Hz±0.5Hz, ensuring continuous power supply to critical loads within the islanded area.

[0135] After the fault is cleared, the station-level collaborative grid controller monitors and confirms that the grid voltage and frequency have returned to normal, and then initiates the seamless reconnection process:

[0136] The station-level collaborative grid controller adjusts the output voltage amplitude of each grid unit to ensure that the voltage amplitude difference between the station and the main grid is ≤5%.

[0137] Adjust the output frequency of each grid unit to make the frequency difference between it and the main grid ≤ 0.2Hz, and use a phase tracking algorithm to make the phase difference between the output voltage of each grid unit and the main grid voltage ≤ 10°.

[0138] Once the voltage amplitude, frequency, and phase all meet the synchronization conditions, the station-level collaborative grid controller issues a grid connection command, and the grid connection switches of each grid unit are closed, achieving smooth grid connection without impact. During the grid connection process, the peak value of the inrush current does not exceed 1.2 times the rated current.

[0139] In some alternative embodiments, the energy storage battery is a lithium iron phosphate battery, the bidirectional converter adopts a three-level topology, and the switching device is an insulated gate bipolar transistor; the energy storage battery is replaced with a lithium titanate battery, the bidirectional converter is replaced with a two-level three-phase voltage source converter with a full-bridge topology, and the switching device is replaced with a silicon carbide metal oxide semiconductor field-effect transistor.

[0140] The lithium titanate battery has a rated capacity of 10kWh, a rated voltage of 50V, a charge / discharge rate of 2C, and a state of charge range of 5% to 95%, offering better cycle life and fast charge / discharge performance. The two-level bidirectional converter with a full-bridge topology has a rated power of 12kW and a switching frequency of 20kHz. The silicon carbide metal oxide semiconductor field-effect transistor has lower switching losses and better high-temperature resistance.

[0141] The implementation logic of each step is the same as in the above embodiment; only the relevant parameters need to be adjusted.

[0142] In the calculation of theoretical output power, no parameter adjustment is required because the battery type does not affect the theoretical output power of the photovoltaic string.

[0143] The discharge power command for the energy storage battery can be adjusted to a maximum discharge power of 20kW based on the 2C charge / discharge rate of the lithium titanate battery.

[0144] In the voltage source control strategy of the bidirectional converter, the switching frequency is adjusted to 20kHz, and the PI regulator parameters are optimized accordingly to ensure voltage and frequency control accuracy.

[0145] The system can still achieve photovoltaic power fluctuation smoothing, string AC bus stabilization and grid support functions, and has certain advantages in terms of fast response and long life.

[0146] In some optional embodiments, based on the above embodiments, the power coordinated scheduling model in step five is optimized by introducing environmental prediction information, adjusting the objective function and constraints of the scheduling model, and improving the foresight and rationality of the scheduling.

[0147] The solar irradiance prediction system installed in the photovoltaic power station obtains the predicted value for the next hour. And input it into the station-level collaborative network controller.

[0148] The objective function of the scheduling model is:

[0149]

[0150] in To add weighting coefficients, This is the theoretical output power prediction calculated based on the predicted irradiance. This represents the predicted total output power.

[0151] Add the following to the constraints: This means limiting the future discharge power of the energy storage battery based on the current SOC and predicted irradiance to avoid over-discharge.

[0152] When performing power coordination scheduling, the station-level collaborative grid controller not only considers the current power balance and SOC balance, but also predicts future power changes in advance, rationally allocates the power support tasks of each controllable unit, reduces the frequent adjustment of the station-level central energy storage system, and improves the system's operating efficiency and stability.

[0153] This embodiment utilizes a hierarchical control method combining string-level local compensation and station-level collaborative scheduling to effectively mitigate power fluctuations in photovoltaic strings caused by local shading, bringing the output power of the string-level self-balancing grid unit closer to the theoretical output power and improving the overall power generation efficiency of the photovoltaic array. Through the grid configuration mode of the string-level self-balancing grid unit, it autonomously maintains the voltage and frequency stability of the string AC bus, reducing dependence on the external power grid. The station-level collaborative grid controller's power collaborative scheduling and dynamic reactive power allocation reduce the operational burden on the station-level central energy storage system, enhance the photovoltaic power station's proactive support capability for the grid, and improve the stability of voltage and frequency at the grid connection point. Virtual synchronous machine cluster control and protective off-grid and seamless re-grid functions further enhance the grid friendliness and operational reliability of the photovoltaic power station.

[0154] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for graded charge and discharge regulation of energy storage to smooth photovoltaic power fluctuations, characterized in that, The method, applicable to a photovoltaic power station system comprising multiple photovoltaic strings, string-level self-balancing grid-connecting units corresponding to each photovoltaic string, a station-level central grid-connecting energy storage system, and a station-level collaborative grid-connecting controller, includes the following steps: Step 1: Each string-level self-balancing grid unit monitors the actual output power of its corresponding photovoltaic string in real time, and calculates the theoretical output power of the photovoltaic string under the current environment based on the operating environment parameters of the photovoltaic string. The string-level self-balancing grid unit includes an energy storage battery and a bidirectional converter with independent grid control capability. Step 2: Each string-level self-balancing grid unit calculates the power difference between the actual output power and the theoretical output power, and compares the power difference with a preset first power threshold. When the power difference is greater than the first power threshold, the photovoltaic string is determined to be a weak string affected by local shading, and the local compensation mode is activated. Step 3: The string-level self-balancing grid unit in the local compensation mode controls the energy storage battery to discharge through the bidirectional converter, so that the total active power output by the string-level self-balancing grid unit approaches the theoretical output power. At the same time, the bidirectional converter of the string-level self-balancing grid unit operates in grid mode to maintain the voltage and frequency stability of the string AC bus connected to the string-level self-balancing grid unit. Step four: The station-level collaborative grid controller collects the operating status, power difference, and state of charge information of all string-level self-balancing grid units through the communication network, and monitors the voltage and frequency of the photovoltaic power station grid connection point. Step 5: When the state of charge (SOC) of the energy storage battery of any string-level self-balancing grid unit in local compensation mode is lower than the preset SOC threshold, the station-level collaborative grid controller initiates power collaborative scheduling. The power collaborative scheduling includes instructing the station-level central grid energy storage system to inject power into the string AC bus, or instructing other string-level self-balancing grid units with an SOC higher than the SOC threshold to adjust the output power of the other string-level self-balancing grid units, so as to provide power support to the string-level self-balancing grid units with low SOC through the string AC bus. Step 6: When the voltage at the grid connection point is detected to deviate from the reference voltage value set for the grid connection point, the station-level collaborative grid controller calculates the total reactive power required for the entire station based on the voltage deviation, and dynamically distributes reactive power adjustment commands to all string-level self-balancing grid units and the station-level central grid energy storage system, so that each string-level self-balancing grid unit can coordinately generate or absorb reactive power while completing its own active power adjustment task, so as to support the stability of the grid connection point voltage. The power coordinated scheduling in step five specifically includes: the station-level coordinated grid controller establishing a scheduling model with the optimization objectives of overall station power balance and balanced state of charge of each energy storage battery, and solving for the optimal power command of each controllable unit. The objective function of the scheduling model is based on minimizing the output power change rate of the station-level central grid energy storage system and maximizing the consistency of the state of charge of each energy storage battery.

2. The method for graded charge and discharge regulation of energy storage to smooth photovoltaic power fluctuations according to claim 1, characterized in that, In step two, the algorithms for calculating the power difference and determining the weak string groups are executed by the local controller embedded in each string-level self-balancing network unit. The theoretical output power... The calculation formula is: in, This represents the theoretical output power of the photovoltaic string under the current operating environment. This indicates the photoelectric conversion efficiency of a photovoltaic module under standard test conditions. This represents the total area of ​​all photovoltaic modules in the photovoltaic string. This represents the solar irradiance collected in real time by the irradiance sensor installed on the string. This indicates the real-time temperature of the photovoltaic module's backsheet, as collected by a temperature sensor. This indicates the reference temperature under standard test conditions. This represents the power temperature coefficient of a photovoltaic module.

3. The method for graded charge and discharge regulation of energy storage to smooth photovoltaic power fluctuations according to claim 1, characterized in that, In step three, the string-level self-balancing grid-building unit operates in grid-building mode as follows: the bidirectional converter adopts a voltage source control strategy. The bidirectional converter adjusts the modulation amplitude and phase through its own controller to autonomously establish and maintain the voltage amplitude and frequency reference of the string AC bus, providing voltage and frequency support for other power electronic devices connected to the same bus, without relying on the voltage and frequency signals of the external power grid.

4. The method for graded charge and discharge regulation of energy storage to smooth photovoltaic power fluctuations according to claim 1, characterized in that, In step six, the specific method for the station-level collaborative grid controller to dynamically allocate reactive power adjustment commands is as follows: based on the current apparent power capacity margin and response priority coefficient of each grid unit, namely each group of cascade self-balancing grid units and the station-level central grid energy storage system, the total reactive power demand is allocated proportionally. Among them, units with larger capacity margins are allocated larger reactive power adjustment amounts, and units in idle or light-load states have higher response priorities.

5. The method for graded charge and discharge regulation of energy storage to smooth photovoltaic power fluctuations according to claim 1, characterized in that, The method further includes: Step 7, Virtual Synchronous Machine Cluster Control: The station-level collaborative grid controller uniformly configures virtual rotational inertia parameters and virtual damping parameters for all string-level self-balancing grid units and the bidirectional converters of the station-level central grid energy storage system. When these bidirectional converters respond to changes in grid frequency, they simulate the rotor motion equations of synchronous generators, providing inertial support and damping for the grid. The frequency-active power droop characteristic equation implemented by the virtual synchronous machine control algorithm is as follows: in, This indicates the active power output value that the network unit needs to adjust based on the frequency deviation. Indicates the frequency droop factor. Indicates the reference value of the rated frequency. This represents the actual frequency of the power grid as collected in real time. This represents the set virtual moment of inertia; The network construction unit includes a string-level self-balancing network construction unit and a station-level central network energy storage system.

6. The method for graded charge and discharge regulation of energy storage to smooth photovoltaic power fluctuations according to claim 5, characterized in that, In step seven, all grid-building units, under the unified coordination of the station-level collaborative grid-building controller, adopt consistent virtual rotational inertia parameters and virtual damping parameters, so that the entire photovoltaic power station appears to the outside world as an equivalent virtual synchronous generator with uniform inertia and damping characteristics, participating in the primary frequency regulation of the power grid.

7. The method for graded charge and discharge regulation of energy storage to smooth photovoltaic power fluctuations according to claim 1, characterized in that, The string-level self-balancing network building unit communicates with the station-level collaborative network building controller, as well as with each string-level self-balancing network building unit, via high-speed industrial Ethernet based on time-sensitive networking technology.

8. The method for graded charge and discharge regulation of energy storage to smooth photovoltaic power fluctuations according to claim 1, characterized in that, The energy storage battery is a lithium iron phosphate battery or a lithium titanate battery, the bidirectional converter is a three-phase voltage source converter with a full-bridge or three-level topology, and the switching device used in the bidirectional converter is an insulated gate bipolar transistor or a silicon carbide metal oxide semiconductor field-effect transistor.

9. The method for graded charge and discharge regulation of energy storage to smooth photovoltaic power fluctuations according to claim 1, characterized in that, The method further includes: Step 8, Protective Off-Grid and Seamless Reconnection: When a serious fault is detected in the string AC bus or the main power grid of the power station, the station-level collaborative grid controller instructs all grid units in the affected area to switch to islanded operation mode. The grid units in this area continue to work together to maintain the voltage and frequency stability of the local power grid. After the fault is cleared and the grid voltage and frequency return to normal, the station-level collaborative grid controller adjusts the output voltage phase and amplitude of each grid unit. The grid units include string-level self-balancing grid units and station-level central grid energy storage systems.

Images